INGE HINTERWALDNER

Parallel Lines as Toolsfor Making Turbulence Visible

Introduction tothe Theoretical Background

Bruno Latour’s idea of the circulating reference. Why is it that we try toachieve new insights, new knowledge, new design by way of making artifactssuch as sketches, diagrams, and models? This ubiquitous practice maysound quite banal, but there was some agitation among scholars when theFrench anthropologist Bruno Latour proposed that the more fabricatedand mediated inscriptions are, the better natural scientists can comprehendreality and the more objectivity can be accumulated. Latour is an intellec-tual who belongs to the ‘‘practical turn’’ in science and technology studies,inquiring into the material manifestations of inscriptions, constructions,and representations. In his seminal early 1990s paper, Circulating Reference,Latour minutely retraced the joint effort of a small interdisciplinary teamconsisting of a botanist, a pedologist, and two geomorphologists in theirmission to find out whether the savanna or the forest was gaining in landarea in the Amazonian province of Roraima (Boa Vista). In order to illus-trate this process for his readers, he created what he calls a ‘‘photophilosoph-ical montage.’’ He distinguished single steps in the research, for example:

1. The scientists inspect a suitable site where the savanna meets the forest.2. They lean over two kinds of maps, pointing at precise locations.3. In the forest, numbered tin tags placed on trees mark former visits.4. The cut specimens of plants are collected, numbered, and put in a certain order

in shelves.5. Later they are inspected on a table.

abstract This article discusses how two physicists—Etienne-Jules Marey and Friedrich Ahlborn—visualized turbulence in air and water around 1900. Their depictions are based upon several creative andconceptual presuppositions that can be revealed by comparing the work of the two, each of whomemployed a field of parallel-aligned lines to depict results. Their similar means of visualizing comparablephenomena turn out to function differently, however, depending on the differences in the ways theselines were conceived and made. Representations 124. Fall 2013 © The Regents of the University ofCalifornia. ISSN 0734-6018, electronic ISSN 1533-855X, pages 1–42. All rights reserved. Direct requestsfor permission to photocopy or reproduce article content to the University of California Press at http://www.ucpressjournals.com/reprintinfo.asp. DOI: 10.1525/rep.2013.124.1.1. 1

And so on. Latour describes the path from the phenomena of the expedi-tion to the shared scientific publication as a cascade of aligned acts ofmeasuring, collecting, mapping, abstracting, selecting, classifying, and num-bering. He is interested in every connection where something is turned intoa notation. Although he is scientifically accurate, he remains at a certaindistance from the processes, especially with regard to the pictorial elements.We are not told why the vertical section looks this way and not that; ifconventions, personal bias, or cultural practices play a role; or to whatextent a preliminary sketch was reworked—or how, or why.

In this project Latour is concerned with an old problem in the philos-ophy of science. Via a chain of reference, of discontinuous and thus ‘‘risky’’intermediate steps, he wants to replace a dualistic representational model(the ‘‘Great Divide’’), one that cannot convincingly bridge the distancebetween a phenomenon and its iconic or textual representation—that is,between world and word. According to Latour, empiricism shows that a ref-erence is established with each work step where small gaps come aboutwhile points of contact are ensured. No single artifact, only the chain ofreference as a whole, is able to carry the burden of representation. Thereference circulates along the chain if each element is available for review.‘‘An essential property of this chain is that it must remain reversible. Thesuccession of stages must be traceable, allowing for travel in both direc-tions.’’1 The symmetry inherent in Latour’s reference system is plausiblefrom a retrospective point of view that mainly looks at the actants. Latourhimself offers some decisive hints that the situation does not really functionas symmetrically as the expression ‘‘reversibility’’ may suggest. He writes:‘‘We see only an unbroken series of well-nested elements, each of whichplays the role of sign for the previous one and of thing for the succeedingone.’’2 The elements are treated in retrospect as signs (or form), and inprospect as thing (or matter). This directionality is crucial. When theresearcher points in the direction of ‘‘thing,’’ and thus tries to think intothe future toward what still has to be found, and when s/he tries to act whilethe next small step is not yet accomplished, the situation proves consider-ably more complex and opaque. If, according to Latour, the representationlies solely in the chain of reference and not in a single picture—and muchspeaks for this hypothesis—how can a scientist then decide about the nextsteps to take? In the end, one has only the ingredients at one’s disposal forproblem solving, but no recipe for finding one’s way through, says themodel theoretician Marcel Boumans.3 The tinkering is carried out throughthe inclusion of the most obvious, namely, what one, with great effort andcontrol, has put in front of one’s eyes and into one’s hands. It is necessary,then, to take a closer look at what is at hand and to investigate how it isavailable.

2 Representations

On the operativity of images. The idea of operativity is not absent in Latour’sthinking, but in my reworking of Latour’s diagram of circulating reference(fig. 1), I have added the function of operativity in order to emphasize itsimportance. The actor-network theory he developed with Michel Callon andJohn Law indeed offers the idea of spaces of change and translations in whichobjects of knowledge are handled not only as information in social systemsbut also as active entities that organize and regulate networks. Thus the agencyof nonhuman entities is emphasized. Nonetheless, Latour is much more con-cerned in his texts with the question of reference, or why scientific outcomescan be so powerful and ‘‘universal’’ though fabricated locally. Latour wouldprobably ask how much operativity (in the sense of ability to connect or tobe reworked) is necessary for the representational chain. I wouldask how muchor which kind of representation is necessary for an image to be operative.

In images and models we find the very particular situation presentedand represented at the same time. Both images and models are in a field oftension between representation (depicting something) and productivity(facilitating something). Science historian Evelyn Fox Keller has named theheuristically separate poles within model theory ‘‘model-of’’ and ‘‘model-for.’’4 ‘‘Operativity’’ can be linked to the latter.

Models and images can be regarded as instruments or as representationsin this context because they are distinguished by the fact that one works withthem and on them at the same time. When analyzing how they are usedoperatively, one pays attention to how they open the scope of activity, providean opportunity for interaction, support or inhibit certain examinations.5 Thenotion of operativity is rarely used in the context of images, probably because

figure 1. Reworking of Bruno Latour’s diagram of circulating reference in thesciences (1999). The added hand and eye stand for full-body engagementwhile operating with the materials at hand. From Bruno Latour,‘‘Circulating Reference: Sampling the Soil in the Amazon Forest,’’ inPandora’s Hope: Essays on the Reality of Science Studies 73 (Cambridge, MA,1999), fig. 2.24. With kind permission of Bruno Latour.

Parallel Lines as Tools for Making Turbulence Visible 3

they are often not seen as places of actions. The work of philosopher SybilleKramer is an exception.6 She speaks of operativity in the context of diagramsand notational iconicity, meaning manageability and explorability as well asan ability to constitute objects and generate results. For Kramer, diagrams arenot only instruments for visualizing but also fields of experiment. This view ofoperativity may serve to explain how images function in processes of cogni-tion and intervention, how something to be invented can be ‘‘found.’’

The Comparison

By comparing two case studies in which each experimenter uti-lized his own complexes of images, building upon his findings and usingthem as instruments of insight in order to find something new, I hope toprovide evidence for the potential of such an approach, one that focuses onproduction—that is, on how experiments were actually prepared and con-ducted. I discuss how two experimental scientists—the well-known Frenchphysiologist Etienne-Jules Marey and the more obscure German zoologistFriedrich Ahlborn—advanced their efforts to visualize physical turbulences(air and water) at the turn of the twentieth century. How did they engagewith their staged model-turbulences? In their aerodynamic and hydrody-namic experiments, they marked the phenomena, created photographicrecords, and compiled synthesizing diagrams. At first the visualizations theydeveloped may seem quite similar, as they both used fields of parallel-aligned lines to represent results. However, any depiction of turbulent flowsis based upon several creative as well as conceptual presuppositions. A closerlook at how these lines are made and conceived shows that, for variousreasons, similar means of visualizing comparable phenomena function dif-ferently. By analyzing the visual logic of such a formal arrangement as thefield of parallel lines, I ask which of the various but not arbitrary usages wereadopted, and to what end. I am keen to extract the differences—which inmy argument point to the function of images, their modus operandi—and toget a better understanding of the individual image techniques. I show towhat extent the adopted techniques and the purpose-made pictorial arti-facts foster or inhibit actions and thus shape the research process. Why, forexample, does one of the scientists seem to stall with his line managementwhile the other experiences research-related intellectual flights of fancy?

The protagonists as reflected in the literature. Of the two researchers, Mareyis known as the one who systematized and expanded the method of graph-ical recordings, as one of the pioneers of cinema, and as a forerunner inmotion-capture technologies, perfecting the use of the black screen.7 He is

4 Representations

a widely discussed figure in visual, film, and media studies, as well as in thehistory of science and history of art. With his dynamographic examinationsMarey revolutionized the relation between the traditional visual arts andphotography. Together with the photographer Eadweard Muybridge, he‘‘solved’’ an enigma that puzzled many artists in the nineteenth century,namely, whether a trotting horse at top speed can have, at any givenmoment, all four hooves off the ground.8 The associated picture series ledto controversial discussions among Muybridge’s contemporaries, as thecamera view did not correspond to the human impression and contradictedall traditional modes of representation. One faction (Georges Gueroult,Auguste Rodin, and Robert de la Sizeranne) insisted upon the differencebetween the artistic truth (truth of ensemble) and the scientific truth (truthof detail). The members of the other faction, academic painters like Jean-Louis-Ernest Meissonier and Jean-Baptiste-Edouard Detaille, after the initialshock used the instantaneous photographs quite literally to correct thepositions of horses’ legs in their hippographic artworks. Although Mareyhimself stated clearly that aesthetics was not his field, together with GeorgesDemeny he conceived an artists’ handbook, a chapter of which he titledLocomotion in Man from an Artistic Point of View, which proposed ways artistscould use chronophotography (for example, Marey writes that the mostvisible moments for the eye are the most intense on the photographic platedue to the accumulated exposure times).9 Less concerned with positivistquestions, the philosopher of aesthetics Paul Souriau appreciated Marey’smultiple exposures on a fixed plate not because of any fidelity to nature butbecause with them a completely new visual language for suggesting theimpression of movement could be gained.10 Avant garde figures in the arts,such as Georges Seurat, Edgar Degas, Frantisek Kupka, and Marcel Duchampwere also attracted by this visual effect. The question of how motion should berepresented even led to deep struggles within the Italian futurist group. WhileGiacomo Balla and the Bragaglia brothers followed Marey in depicting suc-cessive phases of a movement, Umberto Boccioni favored the idea of durationand potential force as expressed by the philosopher Henri Bergson—whoworked at the same time and institution, covering the same topics as Marey,but holding an opposite view.

According to Marta Braun, who in her seminal book Picturing Time pro-vides an insightful overview of his rich body of work, Marey’s images becamethe ‘‘dominant twentieth-century pictorial convention of the dynamic sensa-tion of time.’’11 The images seem to correspond to the wish to depict modernexperiences of speed and dynamism, and Marey’s work in turn became ‘‘thekey visual source of this aesthetic modernism.’’12 Thus, it is not surprising thatmany scholars today acknowledge a kind of influence from Marey on youngergenerations of artists and filmmakers and on some new technological

Parallel Lines as Tools for Making Turbulence Visible 5

inventions. Where such a genealogy is not the primary goal, Marey’s work isoften discussed within the realm of theory-laden concepts like ‘‘memory,’’‘‘trace,’’ the poetics of the seemingly automatic ‘‘self-inscription’’ of move-ments, and the ‘‘visualization of the invisible.’’ My approach points more tothe hands-on side of image- or model-based research—in other words, to the‘‘microbricolage’’ with artifacts, to use Francois Dagognet’s expression.13

Relatively few scholars have dealt with Marey’s last experiments, in whichhe revealed himself as a pioneer of wind tunnels.14 The centennial exposi-tion at the Musee d’Orsay, in Paris 2004, entitled Mouvements de l’air wasdevoted to these experiments. The audience was allowed to play with fivevery impressive reconstructed wind tunnels. Marey himself hardly publishedany theoretical commentaries (just four short notes) on his aerodynamicimages.15 Is this fact a symptom of Marey’s being ‘‘a polymorphous andbulimic [boulimique] scientist’’?16 This quite dramatic description may sug-gest simply that Marey did not rest long with one problem and was soonattracted by another topic around his general theme, ‘‘movement.’’

I am more sympathetic to the opinion of theater scholar Daniela Hahn,who says that turbulence also affected Marey’s epistemic practice and there-fore irritated him.17 However, she offers no reason why this might be thecase. The science historian Christoph Hoffmann, who compares Marey’swork with the streamline experiments of Ludwig and Ernst Mach, showsthat quantification became a (largely unsolved) issue in scientific photog-raphy around 1900. In his view, these researchers were ‘‘to a certain extent‘dazzled’ or ‘trapped’ by the capacities of their favorite tools.’’18 I think it isimportant to open this black box of ‘‘tools,’’ or ‘‘media,’’ especially as Mareyand Ahlborn did not just apply an already fully developed technology totheir experiments. The recording devices themselves can be seen as objectsof study. Furthermore, I would like to propose an explanation for Marey’slack of success in this case by pointing to the difficulties he encountered inthe kinds of images he produced and how he (most likely) engaged withthem. His lack of success is closely linked to habits, persistent ways of think-ing, established or personal problem-solving strategies, and so on.

But, in general, Georges Didi-Huberman is correct in saying that Marey wasgifted in constructing open experimental setups that could surprise him timeand again and allowed him to be productive. According to Didi-Huberman,Marey’s imaginative genius lay in the fact that ‘‘he prolonged these surprises ina heuristic sense and instrumentalized them in a novel way without consideringthe rule previously axiomatically defined for the experimental apparatus.’’19

Although Friedrich Ahlborn was, in my opinion, equally versatile andinventive with respect to his imaging techniques, he has remained largelyunknown. Practically no literature exists about him.20 While he was certainlya known member of the German-speaking scientific community at the time

6 Representations

—not least because of his long-running controversy with Ludwig Prandtl,one of the leading flow researchers—he has not left a noticeable trace in theart world. This is no doubt due to the fact that he did not have the means tospread his ideas in richly illustrated volumes. Nonetheless, Ahlborn’simages, like Marey’s, can be examined in aesthetic terms. He put consider-able effort into achieving ideals like clarity by eliminating unwanted arti-facts, cropping, accentuating contrasts, symmetrizing, or finding reducedand concise form(ula)s for complexities. This was likely important for con-vincing himself first of all, and then his colleagues, of the value of his find-ings, even though it meant accepting a certain tension between singularoptical results and reconstructions based on accumulated impressions.

If in the following I address the unquestionably aesthetic images withrespect to their purpose, that is, for their epistemic aspect, it is not because Iam writing the history of the visualization of turbulence. Likewise, my recov-ery of Ahlborn is not motivated by ambition to close a gap in the history ofexperimental physics. Instead, through the following two case studies I wantto discuss by example the roles of images in situations of knowledge pro-duction. The pioneering attempts of Marey and Ahlborn to deal scientificallywith what later would be called chaotic dynamics were carried out in a situa-tion where no secure knowledge and no established procedures yet existed.Another reason their experiments are suitable for my approach is that bothdeveloped a three-step procedure for gaining and reworking their images,making it possible to examine what happens with and between each step.

Similarities/Things in Common. What are the commonalities between thetwo scientists? Both examined birds’ flight and moved their research focusfrom the kinematics of the wing to its effect on the surrounding medium.Thus, initially, both experimented with an air gauge. Soon each droppedthis technology because they realized how easily their object of investigationcould be influenced. Ahlborn wrote, ‘‘The mere measurement is blind.Whoever is able to photograph the current, in addition to the mere mea-surements, can evaluate his results with much greater accuracy.’’21 This iswhy they stopped using measuring apparatuses within the sensitive mediumitself. In any case, both of them hoped to get to an understanding of theinner relations and the distribution of forces by making the invisible visible.Both worked as outsiders with only a few collaborators, and each had finan-cial worries. Both were wholehearted experimenters and struggled with thefact that in most cases they were not able to derive a mathematical formu-lation from their qualitative findings. They felt compelled to put existingflow theories on a reliable foundation through explicit experiments withideal frictionless liquids.22 Around 1900, both turned to the topic of aero-dynamics and designed individual experimental setups that permitted

Parallel Lines as Tools for Making Turbulence Visible 7

optical access to the study of turbulence. Both let the findings from theirpictorial results leave their mark when it came to making adaptations to theexperiments. In their visualization methods, they relied on a sharedsource of inspiration, the pioneer Ludwig Mach.23 Marey became knownthrough chronophotography; Ahlborn occasionally called his pictures‘‘photochronographies.’’ These expressions meant something different ineach case, but for both photography was combined with temporality. Aston-ishingly, Marey also called the photographs of his wind tunnel ‘‘chronopho-tography.’’ If ‘‘chronophotography’’ is defined as a repeatedly exposedphotographic plate, I hope to show why the wind-tunnel photos are notchronophotographs.24

In the experiments with flow, both investigators used the visual markingof the phenomenon, which then appeared as a streak formation in thephotographs. These streaks came about differently (I will come back to thisissue later). In each case, the photographs served as the starting basis fortheir respective analyses, which were then interpreted in hand-drawn sche-matic diagrams. Thus both scientists had developed a three-step procedure:first both established differentiations by marking the phenomenon; second,the photographic record was analyzed; and third, the photographic recordwas synthesized graphically. The following comparison considers the pas-sages between the three steps of the procedures and focuses on a specialformal depiction of phenomena used by both: namely, a field of parallel-aligned lines. This field of lines turns out to be a productive instance ofrepresentation that allowed both scientists to bring clarity into the nonlin-ear dynamic in the first place. It influenced the research process by openingand facilitating some paths of inquiry and by resisting variation in otherrespects. To what extent these lines can be further developed or applied inmultiple ways also depends how they are made and conceived. And here thedifferences begin.

Case Study I:Etienne-Jules Marey’s Wind Tunnels

Between the years 1899 and 1901, Etienne-Jules Marey con-structed four so-called two-dimensional wind tunnels. In his book Le vol desoiseaux (1890) he had already explained, regarding air resistance, that it isirrelevant whether it is the object or the air that moves, as long as one or theother is at rest.25 In his wind tunnels he chose a resting obstacle—which isconsistent with his previous work in which it was always dynamic bodies thatwere examined, in that he now directed his attention to the moving air.Marey’s wind tunnels consist of a vitrine into which smoke is piped from

8 Representations

above through equidistant nozzles. In later versions the smoke filaments aresuctioned out from below. The first challenge of this experimental settingwas the production of a regular air current. In order to eliminate unwantededdies (remous) when introducing the smoke, Marey invented a kind of filtersystem with fine silk gauze stretched over a wooden frame. ‘‘In order toprevent the fan attached to the container below and suctioning down thesmoke filaments from likewise turning into a ‘cause de troubles,’ silk gauzeswere also arranged on the bottom to regulate the airstream.’’26 In this way,subtle smoke streaks develop—arranged like the strings of a lyre, wroteMarey.27 In the approximate middle of the vitrine an object was introducedso that the—here now desired—vortices (tourbillons) of moved air couldunfold. Thanks to the experiments he conducted in an aquarium in 1893,Marey was confident that chronophotography could serve for the study ofair movement as well as it had for the study of water, and it would show howthe air threads behave when meeting obstacles.28 Consciously he chose the‘‘distanced medium’’ of photography in order not to compromise the easilyperturbable air. But Marey struggled, as did Ahlborn, with the shock wavesand the dust formation caused by the photographic flash.29 The flash ‘‘sur-prises’’ the curling streaks, but on the other hand a permanent illuminationheats the air.30 The issue of lighting seems to play a central role in turbu-lence research in general.31 The difficulty with light was probably not thesole reason that Marey made only snapshots of his wind-tunnel experiments,and not photographic series, multiple exposures, or films, as he had done inhis research into reproducing movement.

A photographic test image (fig. 2) has been preserved that shows Marey’sfourth and last wind tunnel without an obstacle. With this he demonstratedthat the undisturbed course of the smoke fills the whole chamber with a reg-ularly striated field. This kind of visual calibration is repeated in each of therecordings in the upper third of the image as a sort of inlet flow. In order togain something scientifically usable, differences must first be established. Inworking with a camera, an optical contrast is necessary. The differencebetween a pointed absence of turbulence in the region ahead of the obstacle,on the one hand, and the event, which then can obtrude distinctly and in fullclarity, on the other, serves to demonstrate the reliability of the setup.

Excursus: Ahlborn’s doubts. In Ahlborn’s view, too, the parallelism of linesindicates ‘‘calm,’’ and the deviation from it, ‘‘event’’ (fig. 3): ‘‘If one were totake a picture without disturbing the water with an immersed body of resis-tance, these lines would necessarily run parallel and all theoretically havethe same length; thus, they would evoke the impression of a uniformlyflowing liquid.’’32 What works in Ahlborn’s water tunnel because of theresting fluid (I will introduce his experimental setup later) is not readily

Parallel Lines as Tools for Making Turbulence Visible 9

transferable to a wind tunnel where air is introduced. He is convinced thathere this calmness is deceptive. Ahlborn indeed appreciated Marey’sadvances. In Ahlborn’s view, however, his colleague (and other contempor-aries) had not yet accurately registered the phenomena because theirrecordings lacked the needed clarity at the crucial location: ‘‘Unfortunately,the perturbances next to the introduced bodies are so vast that the finesmoke filaments often lose their contours as they approach the obstacle.And thus, the most essential part of the appearances remain hidden ina uniform mist. Furthermore, the method has the disadvantage that thevelocity of the flow cannot be varied much. Further consequences have notbeen drawn from these images,’’ at least no correct ones, said Ahlborn.33

This is not necessarily due to the incompetence of the researchers, ‘‘in fact,the immediate observation of this form of air movement [turbulence] isimpossible, not only because air is invisible—this problem could in a pinchbe remedied by introducing floating visible bodies [Sichtkoerper]—but alsobecause it is a matter of highly involuted, spinning, and at the same time

figure 2. Etienne-Jules Marey,photograph of wind tunnel with

57 injectors, without obstacle,1901. From Georges Didi-

Huberman and LaurentMannoni, Mouvements de l’air,

Etienne-Jules Marey, photographe desfluides, (Paris, 2004), plate xii.

Collection Cinemathequefrancaise, Paris.

10 Representations

progressing movement.’’34 According to Ahlborn, air is not suitable forinvestigations because it cannot be controlled to a sufficient degree. Indeed,in the wind tunnels one tries to master the situation by switching in so-calledflow straighteners. These are wide-meshed alveolar grids made from metalsheets, which are often arranged consecutively in a row. In this way theairflow caused by a turbine is brought into regulated pathways. The photo-graphs of figure 4 depict a wind tunnel without a test object. They show theconsequences of staggered, built-in antiturbulence grids on air flow of a con-stant velocity (12.2 meters per second). Obviously, these devices (as well asMarey’s nozzles) achieve the desired effect by divesting the wind of most ofits lateral movement.

Nonetheless, Ahlborn expresses his reservations: ‘‘The flow straightenerhowever is not only a destroyer but also a generator of turbulent movementbecause due to friction new eddies must develop along its planes/surfaces’’(see fig. 5).35 And elsewhere he writes, ‘‘The wind jet obtains, through thedouble layers of opposing, rotating vortices, the same cellular arrangementas the flow straightener and the rotating columns of liquid of the vortices,which envelop each cell space in dense succession, forming a sort of skele-ton in the flow that opposes a certain resistance against the deformationthrough outer forces.’’36 The artificial double-rowed Karman ‘‘vortexstreets’’ (a repeating pattern of swirling vortices) disadvantageously affectthe research results. For Ahlborn, this is reason enough to study the aero-dynamics via the hydrodynamics.

The graphical and chronophotographical methods as precursor techniques. Letme now return to Marey and his smoke filaments, which, at least macro-scopically, run in linear paths. Although attractive, pleasing photographicrepresentations of them were not his aim. In order to better anticipateMarey’s next steps in light of the test arrangement of his wind tunnels it

figure 3. Friedrich Ahlborn, sketchshowing stream lines in even (left)and whirling or turbulent (right)flow. Deutsches Museum Archiv,Bildstelle, Munich.

Parallel Lines as Tools for Making Turbulence Visible 11

is necessary to understand the methods he had developed and practiced upto that time.

It has become standard in Marey scholarship to divide his work into twoor three distinct stages. Dagognet, for example, speaks of ‘‘three Mareys’’:one studying internal organic movement with the graphic method (until1870), the second examining external locomotion with chronophotography(until 1890), and the third dealing primarily with physical ‘‘subjects’’—beyond solid bodies (until the end of his life).37 John Douard objects thatthese distinctions could be somewhat misleading because ‘‘Marey main-tained throughout his career a constant set of theoretical backgroundbeliefs, a commitment to attenuated and accessible visual display (thegraphical method), and an experimental heuristic of simple decomposi-tion.’’38 I also think it is crucial to take into consideration the fact that hecarried the results and techniques of all his research into each new problem.

Marey is generally associated with the ‘‘graphical method.’’ Though he isconsidered the method’s eponym, a status further consolidated with his

figure 4. F. N. M. Brown, plate with results of different alignments (0-, 2-, 5-, and12-grid screens) in wind channel, without obstacle, 1971. From F. N. M.Brown, See the Wind Blow (South Bend, IN, 1971).

12 Representations

book La methode graphique dans les sciences experimentales (1878), he is not itsinventor. The ‘‘graphical method’’ can be traced back to the beginning ofthe seventeenth century.39 In order to obviate or lessen the need for conven-tional vivisection, Marey designed noninvasive ‘‘sensitive automata,’’ whichideally could gather ‘‘autographic’’ traces from an intact body. Bodily pro-cesses like the pulsating expansion of arteries or the movement of the chestwhile breathing were transmitted mechanically to a quill. The course of therecorded dynamic appeared as a continuous white thread on a constantlyrotating, carbon-black cylinder. The recorded horizontal lines simply meantthe passing of time; a deviation, however, marked an event, as, for example,the muscle-jerk of a frog’s shank provoked by an electric shock (fig. 6).

Marey synchronized different simultaneous recordings of an occurrenceon various cylinders by assembling them in a single diagram so that correla-tions could be more easily discerned. But in the wind tunnel, Marey did notconduct direct measurements. Therefore it is useful to recall a second com-ponent that he utilized in his aerodynamic studies: the photographicrecording technique. Marey regarded it merely as a special case of thegraphical method, to the extent that ‘‘it allowed for the ‘inscription demouvements extremement rapides’ as well as for registering movements

figure 5. Friedrich Ahlborn,‘‘vortex streets’’ behind a grid(flow straightener), 1931. Theplate’s length is 10 cm; thedistance between the trails is 2cm. The white lines are due toa light reflection accompanyingthe rear end of the singleelements of the flowstraightener. From FriedrichAhlborn, ‘‘Uber theoretischeund naturliche Stromungen,’’Flugwesen 10, no. 1 (1930): 5,fig. 8.

Parallel Lines as Tools for Making Turbulence Visible 13

that do not have enough motor power to guide a quill over the paper.’’40

The graphical method also has the disadvantage of slightly distorting therecorded curves due to the friction and inertia of the links in the machine’schain. Marey found a solution for this problem with the help of a system thathas no inertia: the light beam.41 With fixed-plate chronophotography hehad several possibilities for capturing a movement. These possibilities indi-cate that the trade-off for visibility is avoidance of the movement constitutedby duration and the introduction of the idea of time into the image.42

If he wanted to keep the spatial features and respect the bodilyappearance, he got continuous blurring. If he wanted to refine the analysisthrough the augmentation of the exposure frequencies, he got multipleoverlappings of distinct positions. In order to differentiate the positions,he had to reduce the body to reference points.43 With the reference points,however, he again caused a kind of uncertainty in need of further interpre-tation. From another perspective, one could say that the multiple exposuremethod adds the idea of something continuous to the snapshot, while it isnot until the curve that is derived thereby appears that discontinuous mea-surement points, placed sequentially side by side, are connected.44 From thechronophotographs of locomotion, Marey extracts graphics that point moreto the movement than to the moved body.

The transcription of the visual traces of chronophotography try to retaintwo opposed, but for Marey equally important, aspects: the ‘‘graphicalmethod’’ sets the bodies’ outer appearance aside in favor of a continuousrecording of the movement, while the ‘‘photographical method’’ renouncesa continuous recording in order to show the body in movement in a discon-tinuous view.

Transfer of the graphical and the chronophotographical methods into the windtunnel experiments. For the wind-tunnel images Marey used the continuous

figure 6. Etienne-Jules Marey, sketchof the functioning of a myograph, 1878.

From Etienne-Jules Marey, La methodegraphique dans les sciences experimentales etprincipalement en physiologie et en medecine

(Paris, 1878), 194, fig. 97. Courtesy ofEditions Elsevier Masson SAS.

14 Representations

line from the graphical method and the multidimensionality from thechronophotographical method (fig. 7). By ‘‘multidimensionality’’ I meanthe consideration of synchronous data from several spatially distributed‘‘measurement points.’’ This is why there is not just one smoke filament,but several (and with each new wind tunnel Marey increased them in num-ber). I am convinced that Marey intended to benefit from the same transferof diagrammatic transcriptions that he had found advantageous withchronophotography.

For this he prepared the object of study in a comparable manner: themovement-carrying elements are successively spread over the whole area ofthe wind tunnel; they are optically amplified; and the rest of the tunnel isdarkened, eliminating spatial depth, limiting the scope of what can happenin the experimental setup, and integrating the measure reference. More-over, if the analytical evaluation always happens by means of the ex postfacto reconstructed trajectories, here, with the smoke filaments, Mareyseems to have succeeded in regaining something from the apparent autom-atism of the graphical recording. The uninterrupted lines are already thereas a motif. With this solution, is Marey nearer to his aim of analytic synthesis?Obviously not. Snapshots like the ones we know from his archive still do notallow the researcher to fix the geometry of the temporal process on paper.Instead, perplexity holds sway. Marey confessed to Samuel Langley, whoorganized a grant to help Marey develop his last wind tunnel, that he wassometimes quite stuck when it came to interpreting certain experimentalresults. This holds, for example, also for the bizarre paths of smokethreads.45 Marey felt himself compelled to the unusual step of begging hiscolleagues for advice: ‘‘I would like to attract the attention of my colleagues,especially of those who study the questions of mechanics, and to ask them toinvestigate the best conditions for obtaining reliable images of how the airbehaves when it comes into contact with solid bodies of various forms.Furthermore, I ask them to enlighten me with respect to the mechanicalinterpretation of the figures that represent nothing other than the kine-matic facts of the problem I am trying to solve.’’46

The field in movement: difficulties and approaches to solving the problem. In thefollowing I would like to briefly sketch why the functional logic of thegraphical as well as the chronophotographical method each fall short forthe wind-tunnel experiments.

(a) The logic of the graph. Marey mentions musical notation as a rolemodel for the development of graphical writing in physiology.47 It is inter-esting that he established this genealogical connection because, besides theuniversal claim accompanying it, he revealed something about his concep-tion of the logic of graphical recording. The notation lays down a sequence

Parallel Lines as Tools for Making Turbulence Visible 15

along parallel horizontal lines (which as pentagrams, however, form rigidreferences for the tone pitch). With respect to movement, the graph func-tions analogously so that one could say that the present event is exactlywhere the quill touches the spool. The more we move away from the posi-tion of the quill and look at the marks left behind, the more we look backinto the past onto former occurrences (or sounds). If we isolate a smokefilament and compare it with a graph, we come to an essential differenceand thus to a problem. While the graph fixes (saves), the streak in the windtunnel remains in motion everywhere, so there is no saved space for the past.Where would the equivalent of the events at the cutting edge of the pres-ent be; where is the newest mark showing what is happening right now, ifthe part of each smoke filament that has proceeded the farthest forward isthe oldest? Here, the vanguard has accumulated the most ‘‘history.’’ Onone hand, the foremost part is the oldest, and if it reaches the lower end ofthe tunnel, it disappears. On the other hand, nothing revealing has hap-pened yet where the smoke streaks freshly enter the scene in the upperpart of the wind tunnel. From these admittedly strange considerations itturns out that the graphical logic of recording does not work here. In anycase, the fixation of events happens differently with the photographicapparatus.

(b) The logic of chronophotography. In his presentation for the World’sFair of 1900 (fig. 8), Marey arranged the wind-tunnel images transversely, sothat the striations lay horizontally. In this way the same horizontal readingdirection as that in the chronophotographical studies is suggested. At firstsight, the difference seems solely to reside in the fact that in the wind tunnelthe movement is not carried out by a clearly defined body at a specific place.However, this homology probably arises more from a wish than from a sim-ilar manageability of the outcome. The main difference lies in the fact thathere the whole field is always filled with movement. For chronophotography,too, Marey could only use the change at the outer edge, because the rest ofthe image surface is reserved for archiving past stages.48 When taking a snap-shot, the constant smoke supply does not disturb the region ahead of theobstacle because the distinction is preserved. As soon as a multiple exposureis envisaged, the constant supply certainly causes multiple superpositions,and nothing is gained. After the obstacle, there is the additional, greaterdifficulty that this visualizing technique loses its marking character by dif-fusing. For the chronophotographical record one needs an area as limitedas possible, in which the most significant events occur followed by darknessso that what is registered does not fade. The narrower this area is, the lessconfusion the events on the photographic plate cause. Hence, an uprightwalking human being is better suited than a horse to this technique, and thetechnique itself proves to be motif dependent.

16 Representations

fig

ur

e7

.C

om

bin

atio

no

fth

egr

aph

ical

and

chro

no

ph

oto

grap

hic

alm

eth

od

sin

Mar

ey’s

win

d-t

un

nel

pic

ture

s.C

oll

ege

de

Fra

nce

.Arc

hiv

es(3

PV

_073

9,3P

V_0

730,

3PV

_008

0);E

dit

ion

sE

lsev

ier

Mas

son

SAS;

Cin

emat

heq

ue

fran

cais

e,P

aris

;E

tien

ne-

Jule

sM

arey

,Geo

rges

Dem

eny

port

antl

eco

stu

me

noi

ra

lign

eset

poin

tsbl

ancs

,in

stan

tan

e18

83,D

epo

td

uC

oll

ege

de

Fra

nce

en19

78,

Mu

see

Mar

ey,

Bea

un

e,F

ran

ce.

Ph

oto

:J.

-C.

Co

uva

l.

If Marey had not been successful in linearizing and in narrowing themovements in his wind tunnel, he would have been forced to give up despitehis enormous repertoire of methods.

(c) The addition of a chronograph. However, it was not Marey’s inclina-tion to surrender too quickly. His further research focused on change in theindicator of movement. In his last wind-tunnel type of 1901, Marey couldchange between two programs: the smoke fillets were either aspirated down-ward in smooth streaks (see fig. 1) or with an undulation (see fig. 9). Thelatter is realized by adding an electrically controlled chronographand slightly shaking the tubes at the rate of ten times per second.Previously Marey had put considerable effort into calming the air in thecontainer of the wind tunnel with silk gauges and nozzles in order to be able

figure 8. Etienne-Jules Marey, ‘‘Chronophotography, Station Physiologique,Physique et Mechanique,’’ 1900, poster for the World’s Fair, MuseeMarey, Beaune, France. Etienne-Jules Marey, Physique et Mecanique,chronophotographies, ensemble monte pour l’Exposition Universellede 1900, Depot du College de France en 1955, Musee Marey, Beaune,France. Photo: J.-D. Lajoux.

18 Representations

to study the development of turbulences under ‘‘sterile,’’ preferably mono-causal conditions. And now this: the purposeful generation of an oscillation!

Science historians are right if they identify the vibrating elementattached to the nozzles as a method of measuring time.49 This inventionof Marey’s speaks to the desire for further cues for analysis and quantifica-tion. Together the smoke fillets form a main direction—the vertical axis—from which deviations to the left and right can be judged easily. If onewanted to optimize the ability to locate the phenomena, one would applya kind of coordinate plane (for example, graph paper), that is, one wouldadd other rows of lines at right angles to the already existing ones. Thesimplest way to get such a grid structure is to interrupt the longitudinalstripes in equal time intervals and thus to place simultaneous dots in a row.In all likelihood, it proved too difficult for Marey to create precisely con-toured smoke dots in the air, however. His ultimate technical solution was toleave the vertical streaks unbroken, so that only the indentations in themmark the time impulses. Nonetheless, with this solution, which includeda ruler placed on one side, Marey was able to gain certain insights intovelocity rates in different regions.

The hypothesis that Marey tried to apply not only a lined pattern but alsoa checked one, and thus to put in disparate markers, takes on some weightwhen we look at his preliminary studies in hydrodynamics (fig. 10). Here herealized some chronophotographs of the movements of waves in water byusing small, silvered, wax and resin balls.50 For the studies of aerodynamicshe had thought of something similar, but due to low air resistance he couldhardly attain the analogously advantageous compactness of the resin pearls.In principle Marey—and after him Ahlborn—realized a punctual marking inwater, both aiming at dotlike entities rather than lines.

Much later, Marey’s idea was perfected, for example by F. A. Schrauband his collaborators (fig. 11).51 In 1965, they published their hydrogen-

figure 9. Etienne-Jules Marey, photograph of windtunnel, with 57 injectors, with shutter function, 1901. FromGeorges Didi-Huberman, Laurent Mannoni, Mouvements del’air, Etienne-Jules Marey, photographe des fluides (Paris, 2004),plate xxix. Collection Cinematheque francaise, Paris.

Parallel Lines as Tools for Making Turbulence Visible 19

bubble method, in which a fine wire, as one end of a DC circuit, is used toelectrolyze water at regular intervals. As a consequence the density of thetracer material is less than the density of the fluid whose motion is madevisible. The German aircraft engineer Alexander Lippisch also seemed tobuild upon Marey’s approach when recording the velocity distributionaround a wing model (fig. 12). He placed nonrecurring, distinct parallelimpulses and then fixed them chronophotographically with four shortexposures. ‘‘Since the puffs at each tube were released at the same time,the line at any station of the test field represents the position of particles ofthe initial parallel flow after a certain lapse of time.’’52 Figure 12 wasdesigned to study the conditions of ascending lift. This kind of depictionprovides acceptably thorough insights when already streamlined modelwings are used, but not in cases where massive turbulences are created.Thus, the resolution in figure 12 would not have been satisfactory forMarey. Moreover he would not have stopped at this stage but would havewanted to attach his proven diagrammatic analysis. For this he neededdifferences along the course of the smoke. Only with a checked rasterencoding the ‘‘equitemporality of the occurrences’’ could Marey havereturned to his proven graphical depiction.53 In this case, he would havediagrammed the links between photographed smoke dots of ‘‘equal age’’with lines in order to achieve a more accurate representation (fig. 13). Suchdiagrams would have made areas of acceleration or deceleration visible.After the transition from the graphical to the (chrono)photographicalmethod, then, Marey would have again needed a more fundamental trans-formation of his rich repertoire of measures in order to adequately deter-mine air movements.

figure 10. Etienne-Jules Marey, chronophotograph showing the movement ofpearls in a liquid meeting a flat obstacle, 1893. From E.-J. Marey, 1830/1904: La Photographie du Mouvement (Paris, 1977), 54. CollectionCinematheque francaise, Paris.

20 Representations

Case Study II:Friedrich Ahlborn’s Water Channels

We now come to the second case study. As I have already men-tioned, Friedrich Ahlborn saw that air posed too many drawbacks asa research substrate, and thus he decided to conduct his experiments—alsothose relating to airplanes—in water basins (fig. 14). In 1901, Ahlborn builthis first water tunnel, at a length of approximately two meters. He mountedrails that ran the length of the basin, upon which a motorized car could run,and used a weight to regulate the traveling speed of the car. A bracketattached to the car was immersed in the basin and served as a mount forthe particular resistance object or obstacle, which was carried through thestill water. The camera could be mounted either to the traveling car (that is,in a fixed relationship to the obstacle) or to the edge of the basin (in a fixed

figure 11. F. A. Schraub et al., photograph and schematic representation of thetime-streak marker technique applied to flow in a contraction, ca.1965. From Wolfgang Merzkirch, Flow Visualization (New York, 1974),43, fig. 2.20 and 2.21. Courtesy of Wolfgang Merzkirch.

Parallel Lines as Tools for Making Turbulence Visible 21

figure 12. Alexander Lippisch, four exposures of a smoke front as it passes overa wing section operating at high lift, ca. 1956. From Henry V. Borst, TheAerodynamics of the Unconventional Air Vehicles of A. Lippisch (Wayne, PA,1980), 1–15, fig. 11. National Air and Space Museum (NASM9A08391), Smithsonian Institution.

figure 13. See fig. 9, reworked by the author. Collection Cinemathequefrancaise, Paris.

relationship to the calm water). In later attempts, he always placed twocameras with identical lenses side by side, so that comparable images couldbe made simultaneously with the static and traveling cameras. Approxi-mately in the middle of its run, the car was to have accelerated to the desiredspeed. At that moment, it closed a circuit via an attachment to its rails,triggering the camera shutter and the flash. Due to this illumination andto the black-painted interior walls of the trough, the de-oiled club mossspores (Lycopodium) that had been strewn on the water’s surface (fig. 15)were set off clearly from the background. If in Marey’s setup the smokefillets serve to make the turbulence visible and constitute a differentiatedfield, in Ahlborn’s water channel the diffused, tiny club moss spores fulfillthis same purpose. The difference is striking. How intensely must Ahlbornhave wished for such an orderly array as the parallel smoke streaks in orderto draw the flow lines. His British colleague Henry Selby Hele-Shawobserved ‘‘strong whirlpool action’’ in his own tubes and found it hopelessto study these to begin with. The wish he expressed with respect to the fluidis revealing and also holds for the tracer material. He wrote that a simplersituation ‘‘would be the case if the water flowed very slowly in a perfectlysmooth and parallel river bed, when the particles would follow one another

figure 14. Friedrich Ahlborn, sketch of experimental setup for photographs ofthe water’s surface with partially submerged obstacles, 1902. FromFriedrich Ahlborn, Uber den Mechanismus des hydrodynamischenWiderstandes (Hamburg, 1902), plate 1, figs. 1–2. L. Friedrichsen &Co. Hamburg.

Parallel Lines as Tools for Making Turbulence Visible 23

in lines called ‘streamlines,’ and the flow would be like the march of adisciplined army, instead of like the movement of a disorderly crowd, inwhich free fights tak[e] place at various points.’’54 But Ahlborn’s club mossspores could not be brought into a regular formation; instead they arrangedthemselves on the water in tiny clumps of varying size—nothing like a homo-geneous distribution. Thus, the whole setup could be considered problem-atic.55 Additionally, the exposure time was not long enough that a singlespore-cluster could have left behind a streak through the whole photo-graphic plate. Instead, Ahlborn was confronted with the next critical hurdle,namely, the production of lines in his drawings by using several recordedsections.

Various photographic modes. When using club moss spores it became clearthat the choice of the exposure time played a crucial role in the streakformation on the photograph. But what did Ahlborn’s recording techniquelook like exactly? For the water surface Ahlborn developed three differentmodes of recording (fig. 16) as applied processes of photographic flowanalysis. These permitted the representation of the movement of fluids inthe form of flow lines, force lines, and path lines. If the model was carriedthrough the water, the static camera with a short exposure time recordedthe system of force lines (the force field), but with a long exposure time itrecorded the path lines. If the camera traveled together with the model,then the traveling camera with a long exposure time recorded the flow lines.‘‘Due to the resting fluid while the photographic camera is moved, theswimming club moss spores produce[d] on the photographic plate a system

figure 15. Friedrich Ahlborn, club moss spores (Lycopodium) on a water surface.Deutsches Museum Archiv, Bildstelle, Munich.

24 Representations

of fine lines—their length depending on the exposure period—throughwhich the direction of currents in the fluid is shown in every detail, withgreatest clarity.56 We name these lines flow lines and regard the obtainedphotographic images as what they seem to be, namely as views of a movementof a stream of liquid against a resting body.’’57

In order to really understand what was to be seen, the researcher hadto mentally assume the position of the camera and combine the type ofmovement with the mode of recording. The photographs from the staticand traveling cameras showed ‘‘so utterly deviating and, at first, surprisingflow images, that for the uninitiated it would seem impossible to demon-strate the correspondence of the events [thus] represented.’’58 Each of themodes of recording offers a different view of the same phenomenon andthus exposes different facets of it. It is not easy to comprehend that one seesalways the same—but differently.

The force lines are to be imagined as a system of rigid lines, which remains in a fixedconnection with the model—the obstacle—no matter whether the latter persists inthe flow or is carried through the fluid. The lines denote the direction of thevelocity at each position of the force field. No liquid particle goes along the forcelines . . .but moves in the force field passing by only momentarily into the directionof the force line by which it is crossed just in that moment and which it has alreadyabandoned in the next instant.59

Thus, the particles travel not along the force lines, but along the flow andpath lines (see fig. 17). The path lines are nothing other than flow linesdeprived of the translational movement of the original parallel flow.

figure 16. Ahlborn’s techniques for photographic imaging of the water’s surface:force lines (left), path lines (middle), and flow lines (right). FromFriedrich Ahlborn, ‘‘Stromung und Widerstand an Platten,’’ Flugwesen10, no. 7 (1930): 77–78.

Parallel Lines as Tools for Making Turbulence Visible 25

figure 17. Friedrich Ahlborn, photographs of a hydrodynamic experiment withtraveling camera creating flow lines (top) and stationary cameracreating path lines (bottom). Deutsches Museum Archiv, Bildstelle,Munich.

While Marey obtained precise lines only with a short exposure time,Ahlborn obtained them with a longer exposure period. Considering thesituation more closely, its complexity becomes apparent because with com-plementary images Ahlborn achieved zones of clarity, but also zones ofblurring. However, the latter were not the product of mixing materials, aswith Marey’s images; they were the result of areas of overlap that werecaused by the recording technique, but which proved to be indispensiblefor making the different zones visible. Ahlborn needed indicators in theshape of points because he could use them to represent movement (whichcan be detected as lines), together with areas in which no movement occurs(which can be detected as points—and not just in the short exposures).60

This means that to have used lines from the outset would have been a dis-advantage in terms of his desire to separate these zones optically. Throughthe splitting of the phenomenon, he created a space for experimentation.Ahlborn used his photographs to give structure to the phenomenon, and heachieved further elucidation through his schematic drawings. With the mag-nifying glass in his hand, Ahlborn studied his photographic plates scrupu-lously, ‘‘What do the photograms teach us about the resistance?—In orderto answer this question, we want to extract the essence from the recordingsin the form of the schematic depiction we obtain if we draw a system ofequally spaced parallel lines in the primary direction through the unaf-fected part of the flow in front of the obstacle and furthermore to allowthis system to follow the photographic lines of flow exactly.’’61 To beginwith, it becomes clear that Ahlborn mentally proceeds from the unaffectedflow to the turbulent zone.

The premise of charting idealized flow lines in identical intervals can beseen as quite presumptuous, because one does not get any hint as to how todetermine the distances between them in the photographs (see fig. 18).And then which of the short section traces are to be followed? How usefula continuous line would be. Instead, Ahlborn has only vectors at his dis-posal, which ‘‘to the eye appear as parts of connected flow lines.’’62 But bydrawing detached lines he creates space so that he is able to overlap differ-ent schemes.

The graphical translation. All structuring procedures can be seen asefforts to make something determinable. Let me recall Ahlborn’s own def-inition of turbulence: ‘‘Under turbulence in general one understands a dis-orderly mess of eddy-like movements, the form, rotational direction, andforce of which are fluctuating and indeterminate.’’63 At first glance, condi-tions in the turbulent zones are not at all revealing. The following descriptionprovides an impression of the events’ temporal component:

Parallel Lines as Tools for Making Turbulence Visible 27

figure 18. Friedrich Ahlborn, West Wind at Heligoland. Flow-line photograph(top) and flow-line diagram (bottom). Deutsches Museum Archiv,Bildstelle, Munich.

The fluctuations in the proceeding of the dragwater’s flows are very conspicuous.. . . Already in the normal positioning [that is, orthogonal to the flow] of the platea nearly perfect symmetry of the vortices is hardly ever given in single moments. Thephotograms always show some, at times quite striking, irregularities. Sometimes theone, then the other vortex branch is more strongly developed, more rounded orprotracted; sometimes they come closer to the plate, sometimes they hang a bitbehind and leave a relatively calm, temporarily nearly stagnant dragwater behindthe plate. Accordingly, the wake is also unstable in the direction and velocity of itsmovement. This is why one mostly sees quite diverse flow images on photographicplates, the recordings of which were exposed shortly in succession twice, and onerealizes how fast these fluctuations take place.64

From this description it is evident that the graphic translations do notmirror any specific situation but are saturated by the researcher’s pool ofexperience. Ahlborn ‘‘cleared’’ the situation for more concision and adjustedthe schemata according to symmetry. He openly admitted that the variouselements of his drawings were made with recourse to several photographs.65

With the transcription of these photographs into line diagrams, Ahlbornsucceeded in exposing the processes. Through the combination of differenttypes of lines from the same scenario, he could find diverse correlations,explain some structural connections, and also give more thorough explana-tions or instructions.66 For Ahlborn the drawings turned out to be highlyproductive. Without going into his scientific results, I focus instead on howAhlborn uses the line. Figure 19 is an example of the combination of variouskinds of lines (flow lines and force lines) representing one situation, which—as mentioned—presuppose different kinds of photographic recordings. Withthis diagrammatic overlay of two different kinds of lines representing thesame situation Ahlborn is able to distinguish discrete areas of activity. Theflow lines are shown as lines throughout the diagram and without arrows. Theforce lines, with arrows, depart more or less in the opposite direction. How-ever, in favor of clarity, Ahlborn refrains from representing the entire forcefield in the area behind the obstacle. Due to the slant of the plate, the‘‘gyromes’’ (Greek for ‘‘rounded,’’ ‘‘swirl’’) turn out different in size, but theyare clearly detectable.67 From their respective centers, some curved dottedlines reach out, suggesting the differences in the water level’s vertical height.Furthermore, one sees the stern wave (W2) behind the vortices, as well as thebow wave (W1) in front of the resistance object. The line A impacts the frontside of the disk where the water separates when flowing around the obstacle,representing the front pressure maximum. This zone should be readablefrom the ideal (drawn) flow lines, as it is identical with the positions wherethese lines are the greatest distance from each other. In order to understandhow the pressure of the water is linked with the distance between the drawnlines it is necessary to consider Ahlborn’s frame of reference.

Parallel Lines as Tools for Making Turbulence Visible 29

Depiction from hydrodynamics and magnetism as models. Ahlborn startedwith concepts completely different from Marey’s. With lines he divided thewater into sections; that is, he likely first chose the interval that would besuitable for observation arbitrarily (as did Marey) and then decided whichamount of water he wanted to conceive as a ‘‘unit.’’68 Of his diagrams Ahl-born wrote, ‘‘Each of the bands lying flat between two neighboring linesthen represents an elementary flow or flow thread, and we can imagine thatthe water flows in it as in a pipe if we disregard the friction.’’69 He referred tomathematician Daniel Bernoulli’s 1738 theorum signifying that in a steadyflow the sum of kinetic and potential energy remains constant. The sameamount of water passes faster and with less pressure through thin tubes thanthrough larger ones (see fig. 20). By analogy, in Ahlborn’s diagrams, diverg-ing flow lines mean less velocity and more pressure; converging lines repre-sent the opposite. ‘‘With the help of this key we are able to read and tounderstand the documents of the flow images.’’70 To be more precise, withthis ‘‘key’’ it is possible to convert the photographs into his schemata. Neitherkind of depiction fully coincides with the other because the following state-ment, for example, holds only for the photographs: ‘‘The length of the lines isthe measure for the flow velocity at any point of the field of resistance.’’71

Between the distance covered by the photographed club moss spores ina known exposure time interval on the one hand and the distance betweenthe drawn flow lines on the other, there exists a negative correlation: thelonger the club moss spores’ traces, the narrower the distance between the

figure 19. Friedrich Ahlborn, combination of flow lines, force lines, water-leveldata, and bow and stern waves, 1904. From Ahlborn, Uber denMechanismus des hydrodynamischen Widerstandes, plate 3, fig. 8. L.Friedrichsen & Co, Hamburg.

30 Representations

drawn lines. In one instance the velocity is given in the logic of cameratechnology, in the other in the logic of the traditional custom of depiction.

Besides Bernoulli, Ahlborn referred to other sources to ground hisconceptualization of flow and its graphical representation. Again and againhe took up the representations of ideal potential flow by Leonhard Euler. Afurther point of reference was the work of the British naval engineer WilliamFroude. Because of their analogue qualities and the geometric similarity,Ahlborn relabeled the ‘‘absolute flow lines’’ ‘‘force lines.’’72 Consistent withhis rejection of the traditional (because misleading) nomenclature, he bor-rows his term from the field of electromagnetics. The work of the Englishphysicist Michael Faraday, especially, turned out to be influential, as Faradayisolated structures by spreading iron fillings in a magnetic field (therebyfollowing Peter Barlow and others).

According to science historian David Gooding, after 1850, when Faradayhad already developed his field theory, he started elaborating and defend-ing this ‘‘theory of lines’’: ‘‘Lines emerged from these patterns, partlybecause lines were already available and a proven means of structuringsensory experience, and partly because of Faraday’s particular interest inthe explanatory potential of a link between vibrations and ‘striations.’’’73

Faraday regarded lines as a necessary method of differential observation fordetermining differences of the magnetic influence on various (para- anddiamagnetic) substances (fig. 21). Here too the deviation of the lines fromperfect parallel alignment coded what was essential, and thus the concep-tual basis of the field as already figured with lines in a differentiated spacewas indispensable. Faraday’s method of observation of magnetic featuresbuilds on the possibility of detecting differences in the magnetic influenceon various substances in a magnetic field:

figure 20. Daniel Bernoulli, water flow in converging(or diverging) tubes, 1738. From Daniel Bernoulli,Hydrodynamica (Strasbourg, 1738), plate 4, fig. 28a.University Library of Basel, shelfmark Jt II 7.

Parallel Lines as Tools for Making Turbulence Visible 31

This method of [differential] observation required lines because Faraday had repre-sented the effect in terms of changes in the density of lines. The ‘‘converging’’ ofmagnetic lines by the ‘‘concentrating action’’ of iron made clear a practical under-standing of the experimental techniques of the 1820s. Faraday now claimed that itwas impossible to interpret his results in any other way. He stated emphatically that‘‘no other method [of representation] could be used with the differential system ofobservation.’’74

The main features are again encoded in the converging of the magneticlines. So, for his flow lines, Ahlborn found in Faraday the idea of the ‘‘dis-tortion of lines’’ already preformed. I restrict myself to mentioning just a fewprecursors in order to suggest that Ahlborn contemplated his line forma-tions with respect to the different intervals they have to each other. For him,the relative width of the pipe bundles was meaningful.

Summary: Toward the RelationshipBetween Recorded and Drawn Lines

In summary, we can say that neither of the two researchers limitedhimself to simply redrawing the lines recorded in photographs. The drawnlines are significant for both. As a starting base, in each case an ensemble of

figure 21. Michael Faraday, simplified depiction of paramagnetic anddiamagnetic matter with respect to ‘‘mere space,’’ 1851. Fig. 1:convergence; fig. 2: divergence. From Michael Faraday, Experimental-Untersuchungen uber Elektricitat [1855], (Berlin, 1891), 3:§2807, fig. 1–2.University Library of Basel, shelfmark Jv IX 32:3.

32 Representations

points has to be arranged in order to get the lines drawn as soon as timebecomes relevant. With his lines, Ahlborn splits the water into differentzones in order first to be able to define structures, second to code velocityand pressure, and third to situate his brilliant narrative descriptions. Withthe narration of the events he could describe the single parts at length, stretchthe time so the processes could unfold in front of the reader’s inner eye. Inhis eloquent description, Ahlborn gave the impression that he could observewhat happened when the obstacle changed its direction in the water. In otherwords, he envisioned an interactive handling, though one has to suspect thatduring the actual experiments—lasting only a few seconds—the immersedresistance object always remained in the same orientation.

Marey too is concerned with the tension between space and time, butdifferently. The relationship between the speed of the movement and theexposure time is a difficult issue in chronophotography on a fixed plate. Ifthe object moves too slowly relative to the recording/exposure frequency ofthe camera, then its outer appearance is ‘‘drowned out’’ by the quantity ofinformation. However, this can still be puzzled out as long as a progression isrecorded. According to Jean-Luc Daval, the segmentation of the movementinto many small spatiotemporal units puts the emphasis on the measure-ment of distance by way of time.75 Bernd Stiegler interprets the focus theother way around: ‘‘All of Marey’s experiments have in common that time isconverted into space and is readable in spatial categories. Pontus Hultendescribes this procedure as ‘transcribing an idea of time in terms ofspace.’’’76 With his spatiotemporal inquiries, Marey ultimately points to thecalculation of the expenditure of energy. The comparison between Marey’svisualizations of flow and those of Ahlborn shows a related approach inmany respects. From their hints regarding requirements and claims one canconclude that both scientists pursued their research in accordance withwhat later will be called ‘‘mechanical objectivity.’’77

They were committed to establishing standard requirements for instru-ments and results, as well as to recording phenomena with as little distur-bance or influence from extraneous factors as possible; or—even better—tolet the phenomena inscribe themselves.78 The provisions for making thingsvisible should solely and clearly reveal the existing situation. Even thoughthe approaches seem to be quite similar at first, upon closer examinationsubstantial differences become apparent.

One significant difference between Marey’s and Ahlborn’s approacheslies in the meaning of longitudinal coding. Or, to put it in other words,Marey thinks along the lines and would have had in mind a portrait of thesuccessive advance of the mobile wall of streaks, while Ahlborn thinks moreacross the lines.79 What must be noted here is the relative width of the tubebundles between the flow lines, which here function as partition markers.

Parallel Lines as Tools for Making Turbulence Visible 33

While the bright striations of the club moss spores in the photographsserve as local velocity indicators, and thus their lengths are informationcarriers, the length of the flow lines in Ahlborn’s drawings does not encodeanything further. The drawn lines fill the whole image in such a way that, forAhlborn, a proceeding ‘‘front’’ is not urgently desired. Therefore, in hisschemes no past is depicted. He composes an average impression of a fullydeveloped movement. In fact, there exist sporadic drawings depictingnascent vortices. However, Ahlborn then presents this genesis in separatedrawn stages, and these graphics are not to be misinterpreted as the repre-sentation of a specific development of eddies. Such ‘‘portraits’’ are rather tobe expected from Marey.

The field of lines cannot be restricted to something merely formal ormaterial. Marey and Ahlborn understood it also within their specific horizonsof rehearsed practices, customs, situating factors, and theoretical frameworks.The images are undoubtedly affected by the scientific culture. To modify W. J.T. Mitchell’s dialectical concept of visual culture: while images can be seen associal constructions, the inverse perspective is also evident, namely, that—inthis case—the sciences are shaped insofar as they are partly based on thevisual and on visualizations.80 The images that result from research in turnshape further research and the design process. This is not always uniformlythe case, but I have attempted to give some indication, by way of example, ofhow and to what extent artifacts of research can foster, inhibit, or channelactions directed toward a scientific outcome.

N o t e s

Benjamin Smith edited this essay for language use. Unless otherwise noted, alltranslations are my own.

The two archives from which most of my research has derived are identifiedhere by the following abbreviations:

PAK Philosophisches Archiv of the University of KonstanzDMA Deutsches Museum Archiv Munich

(Within the DMA, the dossier of Friedrich Ahlborn is referred to as NL091, the dossier of Ernst Mach as NL 174; and the dossier of LudwigMach as NL 179.)

1. Bruno Latour, ‘‘Circulating Reference: Sampling the Soil in the Amazon For-est,’’ in Pandora’s Hope: Essays on the Reality of Science Studies (Cambridge, MA,1999), 24–79, here 69.

2. Ibid., 56.

34 Representations

3. Marcel Boumans, ‘‘Built-in-Justification,’’ in Models as Mediators: Perspectives onNatural and Social Science, ed. Mary S. Morgan and Margaret Morrison (Cam-bridge, 1999), 66–96; here 91–93.

4. Evelyn Fox Keller, ‘‘Models Of and Models For: Theory and Practice in Con-temporary Biology,’’ in ‘‘Proceedings of the 1998 Biennial Meetings of thePhilosophy of Science Association. Part II: Symposia Papers,’’ supplement, Phi-losophy of Science 67 (September 2000): S72–S86.

5. A strong statement for operativity is given by William Wimsatt, who explains towhat extent false models can nonetheless be useful. By calling them ‘‘falsemodels,’’ or theories, he means that they lack realism and fail as correct descrip-tions of the world. If they are nonetheless useful, this must be independent ofor at least not solely a product of how well they represent the world. The falsityof a model is often essential to its role. They are false because incomplete,idealized, and only locally applicable, and they misdescribe the interactionsor presuppose a number of entities that do not exist. After listing to what extentmodels can be false, Wimsatt nonethess adds twelve ways of applying themgainfully. In short, the function of representation and the function of opera-tivity can diverge in a single model. See William C. Wimsatt, ‘‘False Models asMeans to Truer Theories,’’ in Neutral Models in Biology, ed. Matthew Nitecki andAntoni Hoffman (New York, 1987), 23–55. William C. Wimsatt, ‘‘Using FalseModels to Elaborate Constraints on Processes: Blending Inheritance in Organicand Cultural Evolution,’’ Philosophy of Science 69, no. S3 (September 2002): S12–S24.

6. See Sybille Kramer, ‘‘Operative Bildlichkeit. Von der ‘Grammatologie’ zu einer‘Diagrammatologie’? Reflexionen uber erkennendes ‘Sehen,’’’ in Logik des Bild-lichen. Zur Kritik der ikonischen Vernunft, ed. Martina Heßler and Dieter Mersch(Bielefeld, 2009), 94–122.

7. On graphical recordings see Robert Brain, ‘‘The Graphic Method: Inscription,Visualization, and Measurement in Nineteenth-Century Science and Culture’’(PhD diss., University of California, Los Angeles, 1996). Joel Snyder, ‘‘Visuali-zation and Visibility,’’ in Picturing Science, Producing Art, ed. Peter Galison andCaroline A. Jones (New York, 1998), 379–97. On the running pictures see, forexample, Marey/Muybridge, pionniers du cinema: rencontre Beaune/Stanford; actesdu colloque, 19 May 1995, Palais des Congres, Beaune (Beaune, 1996). Pierre-Jean Borey compares four aspects of technical problems that Marey trans-formed in his revolving-disk cameras with four main types of filmic montage;see Pierre-Jean Borey: ‘‘Vie et mort dans l’image, de Marey a Marker,’’ inMarey, Penser le Mouvement, ed. Christian Salomon (Paris, 2008), 113–39. Onmotion-capture technologies see Edwin Carels, ‘‘Biometry and Antibodies:Modernizing Animation/Animating Modernity,’’ in Animism, ed. AnselmFranke (Berlin, 2010), 1:57–74. Eric S. Faden, ‘‘Chronophotography and theDigital Image: Whoa . . . Deja vu!’’ in Arret sur image, fragmentation du temps/StopMotion, Fragmentation of Time, ed. Francois Albera, Marta Braun, and AndreGaudreault (Lausanne, 2002), 335–45. The art historian Noam M. Elcott iscurrently working on the black screen.

8. See Francoise Forster-Hahn, ‘‘Marey, Muybridge, and Meissonier: The Studyof Movement in Science and Art,’’ in Eadweard Muybridge: The Stanford Years,1872–1882 (Stanford, 1972), 85–109 (Stanford University Museum of Art 7.10.–3.12.1972; E. B. Crocker Art Gallery, Sacramento 16.12.1972–14.1.1973:University Galleries, University of Southern California, Los Angeles 8.2.–11.3.1973).

Parallel Lines as Tools for Making Turbulence Visible 35

9. Etienne-Jules Marey and Georges Demeny, Etudes de physiologie artistique faites aumoyen de la chronophotographie. Premiere Serie, vol. 1, De mouvement de l’homme (Paris,1893).

10. Paul Souriau, La suggestion dans l’art (Paris, 1893).11. Marta Braun, Picturing Time: The Work of Etienne-Jules Marey (1830–1904) (Chi-

cago, 1992), 264–65.12. Ibid., 264.13. Francois Dagognet, Philosophie de l’image (Paris, 1984), 75.14. An early account was written by Pierre Nogues, Recherches experimentales de Marey

sur le Mouvement dans l’air et dans l’eau (Paris, 1933), esp. 81–97.15. See Etienne-Jules Marey, ‘‘Des mouvements de l’air lorsqu’il rencontre des

surfaces de differentes formes,’’ Comptes Rendus des Seances de l’Academie desSciences 131 (July 16, 1900): 160–63. Etienne-Jules Marey, ‘‘Changements dedirection et de vitesse d’un courant d’air qui rencontre des corps de formesdiverses,’’ Comptes Rendus des Seances de l’Academie des Sciences 132 (June 3, 1901):1291–96. Etienne-Jules Marey, ‘‘Les mouvements de l’air etudies par la chrono-photographie,’’ La Nature (September 7, 1901): 232–34. Etienne-Jules Marey,‘‘Le mouvement de l’air etudie par la chronophotographie,’’ Journal de PhysiqueTheorique et Appliquee, 4th ser., 1 (1902): 129–35.

16. Laurent Mannoni, ‘‘Marey Aeronaute: De la methode graphique a la souffle-rie aerodynamique,’’ in Georges Didi-Huberman and Laurent Mannoni, Mou-vements de l’air, Etienne-Jules Marey, photographe des fluides (Paris, 2004), 5–86,here 53.

17. Hahn wrote a valuable contribution to the turbulence studies of Marey byincluding French philosophy (Michel Serres, Gaston Bachelard); see DanielaHahn, ‘‘Tourbillons et turbulences. Zu einer Asthetik des Experiments inEtienne-Jules Mareys Machines a fumee,’’ ilinx 1, no. 1 (2009): 43–69.

18. Christoph Hoffmann, ‘‘Superpositions: Ludwig Mach and Etienne-Jules Mar-ey’s Studies in Streamline Photography,’’ Studies in History and Philosophy ofScience 44, no. 1 (2013): 1–11, here 10. Independently and in ignorance of oneanother’s research, in our simultaneously written articles we came to similarconclusions with respect to Marey.

19. See Georges Didi-Huberman, ‘‘Der Strich, die Strahne,’’ in Offnungen. Zur The-orie und Geschichte der Zeichnung, ed. Friedrich Teja Bach and Wolfram Pichler(Munich, 2009), 285–99; here 287.

20. Biographical considerations can be found in Johannes Weissinger, ‘‘FriedrichChristian Georg Ahlborn,’’ in Neue Deutsche Biographie, ed. Historische Kommis-sion bei der Bayerischen Akademie der Wissenschaften (Berlin, 1953), 1:107–8.J. Georgi, ‘‘Professor Dr. Fritz Ahlborn, ein vergessener Pionier der Stromungs-forschung,’’ Abhandlungen und Verhandlungen des Naturwissenschaftlichen Vereinsin Hamburg n.s., 2 (1957): 5–18. See also Peter Supf, Das Buch der deutschenFluggeschichte (Stuttgart, 1956), 1:221–24. Inge Hinterwaldner, ‘‘Model Buildingwith Wind and Water: Friedrich Ahlborn’s Photo-optical Flow analysis’’(forthcoming).

21. Friedrich Ahlborn, Dickes und dunnes Profil [Ahlborn-Kanal], October 1, 1925,typescript DMA/NL 091, vi–vii.

22. See Etienne-Jules Marey, ‘‘Le mouvement des liquides etudie par la chrono-photographie,’’ Comptes Rendus des Seances de l’Academie des Sciences 116 (January16, 1893): 913–24, here 914. Revealing for Ahlborn’s attitude toward the ‘‘unre-alistic theory’’ are many of his articles’ titles closing with the phrase ‘‘and thereality.’’

36 Representations

23. See Ludwig Mach, [untitled], in Akademischer Anzeiger [der kaiserlichen Akademieder Wissenschaften in Wien] (July 13, 1893): 198–200. Ludwig Mach, ‘‘Uber dieSichtbarmachung von Luftstromlinien,’’ Zeitschrift fur Luftschiffahrt und Physikder Atmosphare 15, no. 6 (1896): 129–39. Besides the citations in the scientificarticles, this link is proven by the personal correspondence. Ahlborn asked,obviously not for the first time, for photographic records from Ludwig Mach:‘‘From the reading of the presentation [Ahlborn’s, included in the letter toMach] you will see how valuable your beautiful photograms are that youpublished with your work about the visualization of air flow lines. Unfortu-nately the images are very scaled down and you yourself mention that in theoriginals one sees more details than in the reproductions. For the accuratecomparison of the flows in the water and in the air, however, it is of majorimportance to inspect all details closely. . . . For this reason I addressed myselfto your father, Professor Dr. E. Mach, with the request to dispose of somephotographs of the flying projectiles that I am especially interested in study-ing, as a loan’’; Friedrich Ahlborn to Ludwig Mach, Hamburg, December 18,1901, DMA/NL 174/0617, 3–4. See the previously mentioned letter: FriedrichAhlborn to Ernst Mach, Hamburg, December 18, 1901, DMA/NL 174/0616.Three months later Ahlborn’s thank-you letter for the receipt of the photo-graphic plates followed: Ahlborn to Ludwig Mach, March 10, 1902, DMA/NL179/023. In the same year, Ahlborn returned the favor by sending some of hisown negatives: Ahlborn to Ludwig Mach, Hamburg, November 14, 1902,DMA/NL 174/0618.

Two letters by Ernst Mach to his son attest to the fact that Marey wasinformed about Ludwig’s work: ‘‘In Paris I was in a meeting of the ParisianAcademy and I found the old acquaintances all together. Marey is very enthu-siastic about the flow lines’’; Ernst Mach to Ludwig Mach, Vienna, July 24, 1896,PAK. Some years later the son broached the subject again, to which Mach senioranswered as follows: ‘‘In the year 1896 in Paris I personally handed a photo-graph of the air flow lines to Marey. Later he also got the treatise. In any case hecompletely forgot the matter because back then he was grieving for the death ofhis mistress. Additionally, he does not speak German and did not read the stuff.What he knows about us is provided to him by [the Russian philosopher andengineer Peter Klementich von] Engelmeyer’’; Ernst Mach to Ludwig Mach,Vienna, December 11, 1900, PAK. In June 1901 Marey reviewed the works ofLudwig Mach: see Marey, ‘‘Changements de direction et de vitesse.’’ See alsoBraun, Picturing Time, 409n38.

24. Marey tries in every possible way to introduce time into images, and he origi-nally uses both terms: ‘‘The change in terminology that Marey undertook in theyear 1887, from photochronography to chronophotography, was based on theawareness—due to the technological differentiation facilitated—of a transitionfrom the ‘chronography’ (the graphical recording) with the means of photog-raphy to the ‘photography of time’ (chronophotography), which has a strategicfocus on time and not on a phenomenon any more. In fact, the chronophoto-graphy on a fixed plate or on film has its authority for all sorts of movement aslong as visible phenomena are present’’; Michel Frizot, ‘‘Notation als graphischeDarstellung und asthetischer Sprung,’’ in Notation. Kalkul und Form in den Kun-sten, ed. Hubertus Amelunxen, Dieter Appelt, and Peter Weibel (Berlin, 2008),55–67, here 61. With the expression ‘‘photochronography,’’ Ahlborn in turnemphasized the velocity measurement that was simultaneously registered on thephotographic plate with sparks: ‘‘I succeeded in visualizing and automatically

Parallel Lines as Tools for Making Turbulence Visible 37

fixing the course of the flows . . . down to the subtlest details on small models bythe way of photographic and chronographic arrangements’’; Friedrich Ahlborn,Vorschlag zur Errichtung eines Hydrodynamischen Instituts Wasserbau fur Schiffahrt inHamburg, Hamburg, July 8, 1903, manuscript, DMA/NL 091, 5. See FriedrichAhlborn, ‘‘Uber den Mechanismus des Widerstandes flussiger Medien,’’ Physika-lische Zeitschrift 3, no. 6 (1901): 120–24, here 121. See also note 50 of this article.

25. Etienne-Jules Marey, Le vol des oiseaux (Paris, 1890), § 133. Ahlborn also leans onthis principle of the equivalence of action and reaction, hence allowing aninversion.

26. Hahn, ‘‘Tourbillons et turbulences,’’ 65.27. Marey, ‘‘Des mouvements de l’air lorsqu’il rencontre des surfaces,’’ 161.28. See Marey, ‘‘Le mouvement des liquides,’’ 923n1.29. See Friedrich Ahlborn, Uber den Mechanismus des hydrodynamischen Widerstandes

(Hamburg, 1902), 30–31. Friedrich Ahlborn, Einrichtung zur Beobachtung undPhotographie der Stromungen an der Unterseite Schwimmender Korper, undated, type-script, DMA/NL 091 h-73, unpaginated.

30. Marey, ‘‘Des mouvements de l’air lorsqu’il rencontre des surfaces,’’ 161.31. It is not only in Thomas Mueller’s discussion of some historical wind channels

that it becomes clear that the whole experimental setup is directed to photo-graphic or filmic recording. Light (something bright, shiny, or self-luminous)and, as a counterpart, a dark background is an essential factor in the experi-ments dealing with turbulence. See Thomas J. Mueller, ‘‘On the HistoricalDevelopment of Apparatus and Techniques for Smoke Visualization of Sub-sonic and Supersonic Flows,’’ AIAA (American Institute of Aeronautics andAstronautics) Meeting Paper, no. 80-0420 (March 1980): 31–44. DOI:10.2514/6.1980-420.

32. Ahlborn, Uber den Mechanismus des hydrodynamischen Widerstandes, 11.33. Friedrich Ahlborn, ‘‘Die Widerstandserscheinungen in flussigen Medien,’’

Illustrierte Aeronautische Mitteilungen. Deutsche Zeitschrift fur Luftschiffahrt 8, no. 6(1904): 185–99 and 231, here 186.

34. Friedrich Ahlborn, Stromungsbilder und ihre Erklarung, undated, 5. Labeled ‘‘Kor-rektur an Herrn Prof. F. Ahlborn,’’ typescript, DMA/NL 091.

35. Friedrich Ahlborn, Ueber den Einfluss der Turbulenz auf Stroemung und Widerstandan Kugeln und Zylindern, undated, typescript, DMA/NL 091 a-2, 2.

36. Friedrich Ahlborn, ‘‘Turbulenz und Mechanismus des Widerstandes an Kugelnund Zylindern,’’ Zeitschrift fur technische Physik 12, no. 10 (1931): 482–91, here483. See also Friedrich Ahlborn, Mangel der Windkanale, undated manuscript,DMA/NL 091.

37. Francois Dagognet, Etienne-Jules Marey: A Passion for the Trace (New York, 1992).38. John W. Douard, ‘‘Etienne-Jules Marey’s Visual Rhetoric and the Graphic

Decomposition of the Body,’’ Studies in the History and Philosophy of Science 26,no. 2 (1995): 175–204, here 190–91.

39. See E. Hoff Hebbel and L. A. Geddes, ‘‘The Beginnings of Graphic Recording,’’Isis 53, no. 3 (1962): 287–324, esp. 293. Soraya de Chadarevian sounds a note ofcaution concerning the transferability of the graphical methods in different(for instance French and German) scientific cultures. See Soraya de Chadar-evian, ‘‘Die ‘Methode der Kurven’ in der Physiologie zwischen 1850 und 1900,’’in Ansichten der Wissenschaftsgeschichte, ed. Michael Hagner (Frankfurt am Main,2001), 161–88.

40. Wolfgang Schaffner, ‘‘Bewegungslinien: Analoge Aufzeichnungsmaschinen,’’in Electric Laokoon. Zeichen und Medien, von der Lochkarte zur Grammatologie,

38 Representations

ed. Michael Franz et al. (Berlin, 2007), 130–45, here 141. For the quotationswithin the quotation, see Etienne-Jules Marey, La methode graphique dans lessciences experimentales et principalement en physiologie et en medecine (Paris, 1878),115 and 648.

41. Laurent Mannoni, ‘‘Die graphische Methode—eine neue Universalsprache,’’ inNotation. Kalkul und Form in den Kunsten, ed. Hubertus Amelunxen, DieterAppelt, and Peter Weibel (Berlin, 2008), 325–30, here 326.

42. See Etienne-Jules Marey, ‘‘La chronophotographie. Nouvelle methode pouranalyser le mouvement dans les sciences physiques et naturelles,’’ Revue generaledes sciences pures et appliquees 2, no. 21 (1891): 689–719, here 690. Etienne-JulesMarey, La chronophotographie (Paris, 1899), esp. 12.

43. Georges Didi-Huberman, ‘‘La danse de toute chose,’’ in Didi-Huberman andMannoni, Mouvements de l’air: Etienne-Jules Marey, photographe des fluides, 173–337,here 213.

44. See Francois Albera, ‘‘Pour une epistemography du montage: le moment-Marey,’’ in Cinema Beyond Film: Media Epistemology in the Modern Era, ed. FrancoisAlbera and Maria Tortajada (Amsterdam, 2010), 31–46; esp. 38.

45. See Etienne-Jules Marey to Samuel Langley, November 19, 1899, SmithsonianInstitution Archives, Office of Secretary, 1887–1907, Record unit 31, book 26,folder 2.

46. Marey, ‘‘Des mouvements de l’air lorsqu’il rencontre des surfaces,’’ 162.47. Etienne-Jules Marey, Du mouvement dans les fonctions de la vie. Lecons faites au

College de France (Paris, 1868), 93. See also Douard, ‘‘Etienne-Jules Marey’sVisual Rhetoric,’’ esp. 178–84.

48. Didi-Huberman names the recorded visual trace of the movement fittingly‘‘time-trail’’; in his opinion we can see the ‘‘wake of the temporalization’’ innateto the chronophotographs; see Didi-Huberman, ‘‘Der Strich, die Strahne,’’295–96.

49. There exist only simple photographs of the smoke images, which, however,Marey does not call ‘‘snapshots’’ but ‘‘chronophotographs.’’ For Laurent Man-noni this nomenclature has nonetheless some plausibility, because all threefacets (‘‘chrono,’’ ‘‘photo,’’ and ‘‘graphy’’) are present in these pictures: ‘‘chrono,thanks to the electric chronograph making the smoke nozzles undulate; photoof course, given that the process demands light (the flash) and the photo-graphic apparatus; graphy because the bright traces left by the smoke tunnelson the sensitive plate are read and analyzed like a graph[ic]’’; Mannoni, ‘‘MareyAeronaute,’’ 59.

50. See Marey, ‘‘Le mouvement des liquides.’’ Etienne-Jules Marey, ‘‘The Experimen-tal Study of the Motion of Fluids,’’ Scientific American 86 (February 1, 1902): 75–76.

51. See F. A. Schraub et al., ‘‘Use of Hydrogen Bubbles for Quantitative Determi-nation of Time-Dependent Velocity Fields in Low-Speed Water Flows,’’ Journalof Basic Engineering 87, no. 2 (June 1, 1965): 429–44.

52. Henry V. Borst, The Aerodynamics of the Unconventional Air Vehicles of A. Lippisch(Wayne, PA, 1980), 1–14.

53. Michel Frizot, ‘‘Les operateurs physiques de Marey et la reversibilitecinematographique,’’ in Stop Motion: Fragmentation of Time, ed. Francois Albera,Marta Braun, and Andre Gaudreault (Lausanne, 2002), 91–102, here 97.

54. Henry Selby Hele-Shaw, ‘‘The Motion of a Perfect Liquid,’’ Nature 60, no. 1558(September 7, 1899): 446–51, here 446.

55. But Ahlborn conceives of the seeming disadvantages in a productive way. Thefact that on the water’s surface small islands of spores alternate with clear regions

Parallel Lines as Tools for Making Turbulence Visible 39

is advantageous for the photographic recording, so much so that Ahlborn him-self boasts of it. If one blows the pellets about 30 cm above the water level into theair, they are distributed very homogenously. This helps the eye see the dynamic atevery location. However, it is not suitable for photography that needs an opticaldifference. Thus Ahlborn disturbs the even distribution by agitating the water,potentially leading to the second problem: that the seeds adhere to each other.But even in this situation, the researcher succeeds, as he thereby learns why hisrivals in Gottingen are misled in their theory about the process of detachment ofthe boundary layer. See Friedrich Ahlborn, Die photographische Stromungsanalyse:Versuchseinrichtungen und Verfahren, undated (after 1917), typescript, DMA/NL091 b-88, 1–7, here 4. A pragmatic approach with available means that allows himto recognize elements in their multiple functionality is typical for Ahlborn:‘‘When I presented the first photograms of resistance flows at the ScientificSociety in Hamburg [Naturwissenschaftlicher Verein zu Hamburg] in June1900, it was rightly asked whether the flow appearances on the surface corre-sponded to those in the inner liquid and whether the former would or would notbe significantly influenced by the surface tension. The answer is that the tensionon the surface of the water is eliminated by the scattered club moss spores as byevery other dust particle’’; Ahlborn, Uber den Mechanismus des hydrodynamischenWiderstandes, 29.

56. Nonetheless, sometimes ambiguities arise that have to do with the specificvisualization: ‘‘In reality many lines turn out to be longer as these optical traceseasily overlap if the clusters of the club moss meal are positioned close in frontof each other. Furthermore, as the chemical process of the flash explosionfades gradually, so the lines also begin to develop by degrees on the darkbackground. With the greatest light intensity they reach their greatest sharp-ness, and on the other end they again trail off in the darkness as they emergedfrom it. Their ends appear most defined in strongly enhanced transparenciesand their projection images’’; ibid., 11.

57. Ibid., 10.58. Friedrich Ahlborn, ‘‘Die Widerstandsvorgange im Wasser an Platten und

Schiffskorpern: Die Entstehung von Wellen,’’ Jahrbuch der SchiffsbautechnischenGesellschaft (1909): 370–436, here 373.

59. Ahlborn, Die photographische Stromungsanalyse.60. In recordings with a long exposure time, the dot-like depictions materialize

only because the relevant area remains in an invariant position with respect tothe camera.

61. Friedrich Ahlborn, ‘‘Hydrodynamische Experimentaluntersuchungen,’’ Jahr-buch der Schiffsbautechnischen Gesellschaft (1904): 417–53, here 423–25.

62. Friedrich Ahlborn, Hydrodynamische Kraftfelder, undated, typescript, DMA/NL091 k-84, 1.

63. Ahlborn, ‘‘Turbulenz und Mechanismus des Widerstandes,’’ 482.64. Ahlborn, Uber den Mechanismus des hydrodynamischen Widerstandes, 26.65. Ahlborn, ‘‘Die Widerstandsvorgange im Wasser,’’ 411–13.66. It remains unclear why Ahlborn most often transcribes only flow and force lines

into his drawings, but hardly ever path lines. As Ahlborn’s photographs aim atrecording striated-blurred as well as dotted zones at the same time, it is striking tosee that in his drawings he most often depicts the situation with lines—regardlessof this given differentiation. Graphically, Ahlborn does not make a distinctionbetween ‘‘moved’’ and ‘‘still.’’ The difference lies merely in the fact that areas thatappear as points are not rendered as tubelike longish formations.

40 Representations

67. The clear contouring of the vortex pair represents a clarification that in natureis nonexistant in the shown form. On another occasion (for the three-dimensional representation), Ahlborn states: ‘‘Needless to say that the . . . sharpcontours chosen for the schematic depiction of these forms do not exist inreality because the circular movement of a vortex gradually fades and dies offtoward the edge and theoretically ceases only in infinity’’; Friedrich Ahlborn,‘‘Turbulenz und Geschwindigkeitsverteilung in Flusslaufen,’’ Physikalische Zeit-schrift 23, no. 3 (February 1, 1922): 57–65, here 59–60.

68. Sometimes in his descriptions, different logics of zoning are revealed. Evenif Ahlborn subdivides the whole scene with his drawn flow lines into waterthreads, this does not stop him from speaking of the ‘‘two’’ lateral flows(these are units encompassing several water threads that do not directlytake part in the eddies). See Ahlborn, ‘‘Die Widerstandsvorgange im Was-ser,’’ 395–96.

69. Ahlborn, ‘‘Hydrodynamische Experimentaluntersuchungen,’’ 423. See alsoAhlborn, ‘‘Die Widerstandserscheinungen in flussigen Medien,’’ 189–90.Daniel Bernoulli, Hydrodynamica (Strasbourg, 1738), esp. chap. 4, §3 and§15.

70. Ahlborn, ‘‘Hydrodynamische Experimentaluntersuchungen,’’ 425.71. Ahlborn, ‘‘Uber den Mechanismus des Widerstandes,’’ 121.72. See Friedrich Ahlborn, ‘‘Der Magnuseffekt in Theorie und Wirklichkeit,’’ Zeit-

schrift fur Flugtechnik und Motor-Luftschifffahrt 20 (1929): 642–53, here 647; Frie-drich Ahlborn, ‘‘Uber theoretische und naturliche Stromungen,’’ Flugwesen 10,no. 1 (1930): 1–6, here 1–2.

73. David Gooding, ‘‘‘Magnetic Curves’ and the Magnetic Field: Experimentationand Representation in the History of a Theory,’’ in The Uses of Experiment: Studiesin the Natural Sciences, ed. David Gooding, Trevor Pinch, and Simon Schaffer(Cambridge, MA, 1989), 183–223, here 209. See also Peter Barlow, An Essay onMagnetic Attractions (London, 1823).

74. Gooding, ‘‘‘Magnetic Curves,’’’ 214–15. See also Michael Faraday, ‘‘MagneticConducting Power,’’ in Experimental Researches in Electricity (London, 1855), 3:200–73, here 203, § 2804.

75. Jean-Luc Daval, La Photographie: Histoire d’un Art (Lausanne, 1982), 72.76. Bernd Stiegler, Theoriegeschichte der Photographie (Munich, 2006), 97–99. For the

quotation within the quotation, see Pontus Hulten, ‘‘Etienne-Jules Marey et lamise a nu de l’espace/temps,’’ in E.-J. Marey, 1830/1904: La Photographie duMouvement (Paris, 1977), 7–8, here 7.

77. See Lorraine Daston and Peter Galison, Objectivity (New York, 2007).78. See Etienne-Jules Marey, ‘‘Necessite de creer une commission internationale

pour l’unification et le controle des instruments inscripteurs physiologiques,’’Journal of Physiology 23, supplement (1898–1899): 6–7. The idea of ‘‘autographicdepictions’’ is expressed in Friedrich Ahlborn, ‘‘Analyse des Widerstandesdurch Stauversuche,’’ in Uber den Mechanismus des hydrodynamischen Widerstandes,42. See also Friedrich Weltzien, ed., Von selbst. Autopoietische Verfahren in derAsthetik des 19. Jahrhunderts (Berlin, 2006).

79. Thirty years after these experiments, the surrealist artist Max Ernst addressedthe pictorial language of the physicists in his series Blind Swimmers of 1934.According to the analyses presented here regarding the use of parallel-aligned lines, one can draw the conclusion that Ernst not only translated Mar-ey’s 1901 aerodynamic studies (as in the version of that painting called BlindSwimmers [Effect of a Touch], 1934, Julien Levy Collection, Bridgewater [CT]; see

Parallel Lines as Tools for Making Turbulence Visible 41

Braun, Picturing Time, 314–16) but also appropriated the logic of lines diamet-rically opposed to Marey’s (thus oriented toward Faraday’s logic; see the paint-ing The Blind Swimmer, 1934, Museum of Modern Art, New York). This confirmsthe findings of Charlotte Stokes, who identified Ernst’s direct sources of inspi-ration in the scientific journal La nature: revue des sciences et de leurs applicationsaux arts et a l’industrie 17, no. 2 (1901). See Charlotte Stokes, ‘‘The ScientificMethods of Max Ernst: His Use of Scientific Subjects from La Nature,’’ ArtBulletin 62, no. 3 (September 1980): 453–65, here 462–463. However, Stokesdoes not give an account of the authors of the adopted images and articles. Inorder to take up the slack, the artist used images from the following articles:Marey, ‘‘Les mouvements de l’air etudies par la chronophotographie,’’ fig. 5;Lucien Bull, ‘‘La photographie des mouvements invisibles: Experiences de M.Hele-Shaw,’’ 247–50, from which Ernst used figs. 3 and 6 or 7. The last twodepictions indeed represent magnetic force lines, according to Bull, Marey’sassistant from 1895–1904.

80. W. J. T. Mitchell, ‘‘Showing Seeing: A Critique of Visual Culture,’’ Journal ofVisual Culture 1, no. 2 (2002): 165–81, here 170–71.

42 Representations